INDEX

Explanations

mentions of governmental or political entities, particularly the federal government

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the term "federal" in relation to government or political contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Twain

-0.90

 Yelp

-0.87

hovah

-0.79

 Coinbase

-0.75

 Bunker

-0.74

 Hoover

-0.74

 firsthand

-0.73

nea

-0.72

 Haley

-0.71

 Lexington

-0.71

POSITIVE LOGITS

 organising

1.21

ised

1.16

 councillor

1.13

 centre

1.13

 colours

1.12

 flavours

1.09

 councillors

1.08

 favourites

1.07

 realise

1.05

 flavour

1.05

Activations Density 0.084%

mentions of governmental or political entities, particularly the federal government

repeated mentions of the term "federal" in relation to government or political contexts

No Comments

No Known Activations